Training server and method for generating a predictive model for controlling an appliance

Abstract

Method and training server for generating a predictive model for the control of an appliance by an environment controller. The predictive model allows a neural network inference engine to infer output(s) based on inputs. The training server receives room characteristic(s), current environmental characteristic value(s), and set point(s) from the environment controller. The training server determines command(s) for controlling the appliance based on the current environmental characteristic value(s), the set point(s) and the room characteristic(s). Each command is executed by the controlled appliance. The training server receives updated environmental characteristic value(s) and determines a reinforcement signal based on the set point(s), the updated environmental characteristic value(s), and a set of rules. The training server executes a neural network training engine to update the predictive model based on: inputs (the current environmental characteristic value(s), the set point(s), and the room characteristic(s)); output(s) (the command(s)); and the reinforcement signal.

What is claimed is: 
     
       1. A method for generating a predictive model of a neural network used for controlling an appliance, the method comprising:
 storing in a memory of a training server a predictive model of a neural network; 
 storing in the memory a set of rules; 
 receiving by a processing unit of the training server at least one room characteristic, the at least one room characteristic comprising a type of activity performed by humans occupying the room; 
 receiving by the processing unit at least one current environmental characteristic value and at least one set point from an environment controller via a communication interface of the training server; 
 determining by the processing unit one or more commands for controlling an appliance based on the at least one current environmental characteristic value, the at least one set point and the at least one room characteristic; 
 transmitting by the processing unit the one or more commands for controlling the appliance to the environment controller via the communication interface; 
 receiving by the processing unit at least one updated environmental characteristic value from the environment controller via the communication interface; 
 determining by the processing unit a value of a reinforcement signal by applying the set of rules to inputs comprising the at least one current environmental characteristic value received from the environment controller before generation and transmission to the environment controller of the one or more commands, the at least one set point, the at least one updated environmental characteristic value received from the environment controller after transmission to the environment controller of the one or more commands and the at least one room characteristic, the application of the set of rules generating the value of the reinforcement signal, the value of the reinforcement signal being one of positive reinforcement or negative reinforcement; and 
 executing by the processing unit a neural network training engine implementing reinforcement training to update the predictive model of the neural network based on:
 inputs comprising the at least one current environmental characteristic value, the at least one set point, and the at least one room characteristic; 
 one or more outputs consisting of the one or more commands; and 
 the value of the reinforcement signal. 
 
 
     
     
       2. The method of  claim 1 , wherein the predictive model of the neural network comprises weights of the neural network, and updating the predictive model comprises updating the weights. 
     
     
       3. The method of  claim 1 , wherein the at least one room characteristic is received from the environment controller via the communication interface of the training server. 
     
     
       4. The method of  claim 1 , wherein the at least one room characteristic further comprises at least one of the following: a room type identifier selected among a plurality of room type identifiers, one or more geometric characteristics of the room, and periods of time when the room is occupied by humans. 
     
     
       5. The method of  claim 1 , wherein the at least one current environmental characteristic value comprises at least one of the following: a current temperature, a current humidity level, a current carbon dioxide (CO2) level, and a current room occupancy. 
     
     
       6. The method of  claim 1 , wherein the at least one updated environmental characteristic value comprises at least one of the following: an updated temperature, an updated humidity level, and an updated carbon dioxide (CO2) level. 
     
     
       7. The method of  claim 1 , wherein the at least one set point comprises at least one of the following: a target temperature, a target humidity level, and a target CO2 level. 
     
     
       8. The method of  claim 1 , wherein the one or more commands for controlling the appliance include at least one of the following: a command for controlling a speed of a fan, a command for controlling a pressure generated by a compressor, and a command for controlling a rate of an airflow through a valve. 
     
     
       9. The method of  claim 1 , wherein the set of rules generates the positive reinforcement or the negative reinforcement in response to an absolute difference between the at least one set point and the at least one updated environmental characteristic value with respect to a threshold. 
     
     
       10. A training server, comprising:
 a communication interface; 
 memory for storing: 
 a predictive model of a neural network; and 
 a set of rules; and 
 
       a processing unit for:
 receiving from an environment controller via the communication interface at least one room characteristic, the at least one room characteristic comprising a type of activity performed by humans occupying the room; 
 receiving from the environment controller via the communication interface at least one current environmental characteristic value and at least one set point; 
 determining one or more commands for controlling an appliance based on the at least one current environmental characteristic value, the at least one set point and the at least one room characteristic; 
 transmitting to the environment controller via the communication interface the one or more commands for controlling the appliance; 
 receiving from the environment controller via the communication interface at least one updated environmental characteristic value; 
 determining a value of a reinforcement signal by applying the set of rules to inputs comprising the at least one current environmental characteristic value received from the environment controller before generation and transmission to the environment controller of the one or more commands, the at least one set point, the at least one updated environmental characteristic value received from the environment controller after transmission to the environment controller of the one or more commands and the at least one room characteristic, the application of the set of rules generating the value of the reinforcement signal, the value of the reinforcement signal being one of positive reinforcement or negative reinforcement; and 
 executing a neural network training engine implementing reinforcement training to update the predictive model of the neural network based on:
 inputs comprising the at least one current environmental characteristic value, the at least one set point, and the at least one room characteristic; 
 one or more outputs consisting of the one or more commands; and 
 the value of the reinforcement signal. 
 
 
     
     
       11. The training server of  claim 10 , wherein the predictive model of the neural network comprises weights of the neural network, and updating the predictive model comprises updating the weights. 
     
     
       12. The training server of  claim 10 , wherein the at least one room characteristic further comprises at least one of the following: a room type identifier selected among a plurality of room type identifiers, one or more geometric characteristics of the room, and periods of time when the room is occupied by humans. 
     
     
       13. The training server of  claim 10 , wherein the at least one current environmental characteristic value comprises at least one of the following: a current temperature, a current humidity level, a current carbon dioxide (CO2) level, and a current room occupancy. 
     
     
       14. The training server of  claim 10 , wherein the at least one updated environmental characteristic value comprises at least one of the following: an updated temperature, an updated humidity level, and an updated carbon dioxide (CO2) level. 
     
     
       15. The training server of  claim 10 , wherein the at least one set point comprises at least one of the following: a target temperature, a target humidity level, and a target CO2 level. 
     
     
       16. The training server of  claim 10 , wherein the one or more commands for controlling the appliance include at least one of the following: a command for controlling a speed of a fan, a command for controlling a pressure generated by a compressor, and a command for controlling a rate of an airflow through a valve. 
     
     
       17. The training server of  claim 10 , wherein the set of rules generates the positive reinforcement or the negative reinforcement in response to an absolute difference between the at least one set point and the at least one updated environmental characteristic value with respect to a threshold. 
     
     
       18. A non-transitory computer program product comprising instructions executable by a processing unit of a training server, the execution of the instructions by the processing unit providing for generating a predictive model of a neural network used for controlling an appliance by:
 storing in a memory of the training server a predictive model of a neural network; 
 storing in the memory a set of rules; 
 receiving by the processing unit at least one room characteristic, the at least one room characteristic comprising a type of activity performed by humans occupying the room; 
 receiving by the processing unit at least one current environmental characteristic value and at least one set point from an environment controller via a communication interface of the training server; 
 determining by the processing unit one or more commands for controlling an appliance based on the at least one current environmental characteristic value, the at least one set point and the at least one room characteristic; 
 transmitting by the processing unit the one or more commands for controlling the appliance to the environment controller via the communication interface; 
 receiving by the processing unit at least one updated environmental characteristic value from the environment controller via the communication interface; 
 determining by the processing unit a value of a reinforcement signal by applying the set of rules to inputs comprising the at least one current environmental characteristic received from the environment controller before generation and transmission to the environment controller of the one or more commands, the at least one set point, the at least one updated environmental characteristic value received from the environment controller after transmission to the environment controller of the one or more commands and the at least one room characteristic, the application of the set of rules generating the value of the reinforcement signal, the value of the reinforcement signal being one of positive reinforcement or negative reinforcement; and 
 executing by the processing unit a neural network training engine implementing reinforcement training to update the predictive model of the neural network based on:
 inputs comprising the at least one current environmental characteristic value, the at least one set point, and the at least one room characteristic; 
 one or more outputs consisting of the one or more commands; and 
 the value of the reinforcement signal.